bakery-ia/services/alert_processor/app/enrichment/business_impact.py

"""
Business impact analyzer for alerts.

Calculates financial impact, affected orders, customer impact, and other
business metrics from event metadata.
"""

from typing import Dict, Any
import structlog

logger = structlog.get_logger()


class BusinessImpactAnalyzer:
    """Analyze business impact from event metadata"""

    def analyze(self, event_type: str, metadata: Dict[str, Any]) -> dict:
        """
        Analyze business impact for an event.

        Returns dict with:
        - financial_impact_eur: Direct financial cost
        - affected_orders: Number of orders impacted
        - affected_customers: List of customer names
        - production_delay_hours: Hours of production delay
        - estimated_revenue_loss_eur: Potential revenue loss
        - customer_impact: high/medium/low
        - waste_risk_kg: Potential waste in kg
        """

        impact = {
            "financial_impact_eur": 0,
            "affected_orders": 0,
            "affected_customers": [],
            "production_delay_hours": 0,
            "estimated_revenue_loss_eur": 0,
            "customer_impact": "low",
            "waste_risk_kg": 0
        }

        # Stock-related impacts
        if "stock" in event_type or "shortage" in event_type:
            impact.update(self._analyze_stock_impact(metadata))

        # Production-related impacts
        elif "production" in event_type or "delay" in event_type or "equipment" in event_type:
            impact.update(self._analyze_production_impact(metadata))

        # Procurement-related impacts
        elif "po_" in event_type or "delivery" in event_type:
            impact.update(self._analyze_procurement_impact(metadata))

        # Quality-related impacts
        elif "quality" in event_type or "expired" in event_type:
            impact.update(self._analyze_quality_impact(metadata))

        return impact

    def _analyze_stock_impact(self, metadata: Dict[str, Any]) -> dict:
        """Analyze impact of stock-related alerts"""
        impact = {}

        # Calculate financial impact
        shortage_amount = metadata.get("shortage_amount", 0)
        unit_cost = metadata.get("unit_cost", 5)  # Default €5/kg
        impact["financial_impact_eur"] = float(shortage_amount) * unit_cost

        # Affected orders from metadata
        impact["affected_orders"] = metadata.get("affected_orders", 0)

        # Customer impact based on affected orders
        if impact["affected_orders"] > 5:
            impact["customer_impact"] = "high"
        elif impact["affected_orders"] > 2:
            impact["customer_impact"] = "medium"

        # Revenue loss (estimated)
        avg_order_value = 50  # €50 per order
        impact["estimated_revenue_loss_eur"] = impact["affected_orders"] * avg_order_value

        return impact

    def _analyze_production_impact(self, metadata: Dict[str, Any]) -> dict:
        """Analyze impact of production-related alerts"""
        impact = {}

        # Delay minutes to hours
        delay_minutes = metadata.get("delay_minutes", 0)
        impact["production_delay_hours"] = round(delay_minutes / 60, 1)

        # Affected orders and customers
        impact["affected_orders"] = metadata.get("affected_orders", 0)

        customer_names = metadata.get("customer_names", [])
        impact["affected_customers"] = customer_names

        # Customer impact based on delay
        if delay_minutes > 120:  # 2+ hours
            impact["customer_impact"] = "high"
        elif delay_minutes > 60:  # 1+ hours
            impact["customer_impact"] = "medium"

        # Financial impact: hourly production cost
        hourly_cost = 100  # €100/hour operational cost
        impact["financial_impact_eur"] = impact["production_delay_hours"] * hourly_cost

        # Revenue loss
        if impact["affected_orders"] > 0:
            avg_order_value = 50
            impact["estimated_revenue_loss_eur"] = impact["affected_orders"] * avg_order_value

        return impact

    def _analyze_procurement_impact(self, metadata: Dict[str, Any]) -> dict:
        """Analyze impact of procurement-related alerts"""
        impact = {}

        # Extract potential_loss_eur from reasoning_data.parameters
        reasoning_data = metadata.get("reasoning_data", {})
        parameters = reasoning_data.get("parameters", {})
        potential_loss_eur = parameters.get("potential_loss_eur")

        # Use potential loss from reasoning as financial impact (what's at risk)
        # Fallback to PO amount only if reasoning data is not available
        if potential_loss_eur is not None:
            impact["financial_impact_eur"] = float(potential_loss_eur)
        else:
            po_amount = metadata.get("po_amount", metadata.get("total_amount", 0))
            impact["financial_impact_eur"] = float(po_amount)

        # Days overdue affects customer impact
        days_overdue = metadata.get("days_overdue", 0)
        if days_overdue > 3:
            impact["customer_impact"] = "high"
        elif days_overdue > 1:
            impact["customer_impact"] = "medium"

        return impact

    def _analyze_quality_impact(self, metadata: Dict[str, Any]) -> dict:
        """Analyze impact of quality-related alerts"""
        impact = {}

        # Expired products
        expired_count = metadata.get("expired_count", 0)
        total_value = metadata.get("total_value", 0)

        impact["financial_impact_eur"] = float(total_value)
        impact["waste_risk_kg"] = metadata.get("total_quantity_kg", 0)

        if expired_count > 5:
            impact["customer_impact"] = "high"
        elif expired_count > 2:
            impact["customer_impact"] = "medium"

        return impact